The Emperor's New Measurement Model

摘要

Humphry's model faces an identification problem that needs to be resolved before the model can be taken seriously as a measurement model. It is ironic that a paper that argues that "the central importance of units ... is all but ignored in psychometrics" should fail to get the scale of the model right. Even if Humphry succeeds in repairing the problem, however, it is unclear in what sense the model goes beyond existing models in a principled way. Since the measurement pretensions associated with the Rasch model remain a subject of so much debate, it seems premature to make such measurement claims about what is nothing less but also nothing more than an extended version of the Rasch model. Although we applaud the attempt to address the question of quantity, too many statistical and conceptual questions are left unaddressed to properly evaluate the worth of Humphry's contribution.